Bayes rule for Covid-19 testing
https://towardsdatascience.com/false-positives-negatives-and-bayes-rule-for-covid-19-testing-750eaba84acd
The article “False positives/negatives and Bayes rule for covid-19 testing” talks about how we use Bayes rule in predicting the confidence of Covid-19 testing. It is necessary and useful because there is no 100% precise testing. Besides testing negative correctly, testing people false negative, false positive, true negative will give them cost, and the lower possibility of testing wrong we have, the lower possibility people will get the cost. That’s why we need Bayes rule to calculate the probability of testing false and testing true. This will let us have a certain degree of belief in our hypothesis. Letting us using previous results to improve our testings.
We can connect this resource to the “Bayes Theorem” topic we studied in class. The case of Covid-19 testing can be classified as “True positive”, “True negative”, “False positive”, and “False negative”. The possibility of each kind of case can be calculated by using Bayes Theorem. What’s more, we can prior knowledge in the form of a Prior probability term in the numerator of the Bayes’ theorem. Where the prevalence rate in the population is low, we can use prescribe subsequent test to increase the confidence in the test results by using chained Bayes computation to improve our testing. The slide of cascades in our class also mentioned covid-19 testing as example of explaining Bayes Theorem.
The Covid-19 is still pandemic today, and testing is a critical daily part in our defense of virus. And the testing accuracy is very important. The Bayes Theorem not only plays an important role in calculating the testing accuracy but also brings us confidence and belief of our testing method. We use Bayes rules to check our testing accuracy and also use it to keep improving it. It’s hard to image that without Bayes rules, how hard it will be to check whether our testing method is reliable.